Plasma Talk 4

Linear Landau damping - The Maths

Just a piece of a vast problem: Energy exchange between waves and particles in a plasma.
Simply put, in terms of the energy transfer direction:
Waves Particles: Particle acceleration, wave damping.
Particles Waves: Wave instability.
The original paper is Ref. [1]. Landau damping is one of the most studied/debated problem
in plasma physics. Nice Maths and Physical derivation
1
.
Calculation overview
Since the calculation is quite subtle and long, it may be useful to get a general overview from
the very beginning. Here are the steps we will follow:
1. Derivation of the dispersion equation (k, ) = 0 from the 1D Vlasov-Poisson system.
2. Landau contour, the continuity requirement and the Laplace transform.
3. Resolution for small damping and any distribution function.
4. Maxwellian distribution.
1 Dispersion Equation
Start from 1D non-relativistic equations
2
for F(x, v = p/m, t) and eld E(x, t),
0 =
F
t
+ v
F
x
e
E
m
F
v
, (1)
E
x
= 4e
_
n
0

_
F(x, v, t)dv
_
, (2)
1
See Kip Thornes Caltech course Applications of Classical Physics, Chapter 21 mostly for the Maths part
at http://www.pma.caltech.edu/Courses/ph136/yr2004/.
2
Easily generalized to 3D.
1
Figure 1: Imaginary part of G =
_
e
u
2
du/u x, for x C. The real axis is a discontinuity.
where n
0
=
_
F
0
dv is the equilibrium density. Assume F = F
0
+ F
1
, with | F
1
|| F
0
|, F
0
being an equilibrium solution. Same for E. The equilibrium electric eld E
0
= 0. Linearizing
Eqs. (1,2), assuming F
1
, E
1
exp(ikx it), gives
0 = iF
1
+ ikvF
1
e
E
1
m
F
0
v
, (3)
ikE
1
= 4e
_
F
1
(x, v, t)dv. (4)
Extract F
1
from the rst equation, and plug it into the second,
(k, ) = 0, with,
(k, ) = 1

2
p
k
2
_
f

0
v /k
dv, (5)
where
2
p
= 4n
0
e
2
/m, f
0
= F
0
/n
0
and f

0
= f
0
/v. This dispersion relation was rst
obtained by Vlasov in 1925 [2]. It shows should be imaginary. Otherwise, we have a
problem, unless f

0
(/k) = 0. The dielectric function (k, ) has therefore a real and an
imaginary part, which for all kind of systems, is related to dissipation.
We could just consider imaginary and take this quadrature as it is, integrating along
the real axis. But theres a problem. The resulting function of is discontinuous, precisely
when crossing the real axis (see Section 2.2). As an illustration, Fig. 1 displays the imaginary
part of G =
_
e
u
2
du/u x, for x C. The discontinuity is obvious around Im(x) = 0. One
part of the plan has to be physically meaningful, and the other not. But which one? We
could try both options, and check that damping comes only when choosing the upper one.
But what if we didnt know in the rst place that a Maxwellian is stable? We shall see that
a Laplace analysis of the problem can fully answer the question, and will indeed tell us that
the physical half-plane is the upper one (Section 2.3).
Admitting for now the upper-plane is the physical one, what do we do with the lower
one? The answer is that we have to analytically continuate the function we have on the
2

/ k
/ k
v
/ k
Figure 2: The Landau integration contour. It is not closed. It lies always on the same side of
the pole. It lies below the pole.
upper-plane, to the lower one. This means nding a function on the lower plane which makes
a continuous, analytical junction, with what we have on the upper one. In this respect, a
unicity theorem from complex analysis helps: if somehow we nd an expression in the lower
plane matching what we have in the upper one, then this is the only one. The Landau contour
is going to do all of that for us: providing a contour of integration equivalent to an integration
over the real axis for Im() > 0, and an analytical continuation of the later in the lower plane
Im() < 0.
2 Landau contour, the continuity requirement and the Laplace
transform
Lets rst give the solution found by Landau, namely the famous Landau contour. Figure
2 shows this integration contour has 3 very distinctive features:
1. The Landau contour is not closed by the usual semi-circle in the lower or upper half-
plane.
2. The pole /k = (
r
+ i)/k must always lie on the same side of the Landau contour.
3. So, which side? The Landau contour goes below the pole.
These contour prescriptions are called the Landau prescriptions, and the corresponding
contour, the Landau contour. We thus rewrite from now on Eq. (5) as
(k, ) = 1

2
p
k
2
_
L
f

0
v /k
dv, (6)
where
_
L
means integration along the Landau contour. Lets now nd out about these 3
features.
2.1 The contour is not closed
The contour just goes from v = to +, and is not closed in the upper or lower half-
plane, as usual, because we have no guarantee f

0
(v) behaves correctly there, so as to cancel
the integration on the semi-circle at innite radius. Indeed, considering a Maxwellian with
f

0
e
v
2
and setting v = R
v
e
iv
to parametrize the integration on a circle of radius R
v
, we
nd f

0
e
R
2
v
cos 2v
which can hardly be considered a vanishing quantity at R
v
for any

v
[0, ] (or [0, ], if you close in the lower half-plane).
3
2.2 Always on the same side
Assume =
r
+ i, with > 0. As long as remains positive in Eq. (5), the calculation
does not pose any conceptual problem as the pole is not on the real axis, and continuity is
guaranteed.
Now, what if approaches 0, and the pole /k even gets to cross the real axis? We would
like (k, ) to be a continuous function of . Assume rst we leave the integration contour
unchanged (the real axis for v), and compare the quadrature for =
r
+i and =
r
i.
The inuence of the pole is mostly felt where the denominator is minimum at v
r
/k, so
lets locally get f

0
out of the integral and compare,
I
1
=
_
dv
v (
r
+ i)/k
and I
2
=
_
dv
v (
r
i)/k
. (7)
The dierence I
1
I
2
is,
I
1
I
2
= 2i
_
/k
(v
r
/k)
2
+ (/k)
2
dv. (8)
The continuity of (k, ) demands the expression above vanishes when 0
+
. The problem
is that it does not. Instead, the quadrature tends to (see function G
2
in Appendix A), so
that we indeed have a jump of amplitude 2i when crossing the real axis
3
.
The only way to avoid this is to deform the integration contour in such a way that it always
lies on the same side of the pole /k.
2.3 The contour goes below the pole
To understand why the contour goes below the pole and not above like in Fig. 4, we need
to follow Landau in rethinking the problem in terms of the time evolution of a perturbation
applied at t = 0. The Fourier technique is not well suited for that because it entails an
integration from t = to +. By design, it does not single out any special moment in
between. By contrast, the Laplace transform involves times only from zero to +. As shall be
checked, the Laplace transform technique gives an unambiguous response about the location
of the pole with respect to the integration contour.
Considering a function h(t), its Laplace transform

h() and the inversion formula
4
, read

h() =
_

0
e
it
h(t)dt, (10)
h(t) =
_
C
L
e
it

h()d, (11)
3
It can also be said that for I1, the integration path makes a counter-clockwise half-turn around the pole,
so that I1 = i. But for I2, the half-turn around the pole is clockwise, so that I2 = i and I1 I2 = 2i.
4
I here follow Landaus book, [3], p. 139, in dening the Laplace transform this way. Thats just the usual
one,
g(p) =
Z

0
g(t)e
pt
dt, (9)
for p = i. It avoids having to rotate everything in the complex plane to relate the calculation to Eq. (5).
4

Re()
Im()
> 0
1
2
3
5
6
4
n
Figure 3: Laplace integration contour. Goes from = + i to ++ i with > 0, and
is closed in the lower half-plane. By design, > 0 and such that every single poles
1
, . . . ,
n
of the integrand lie inside the contour.
where the contour C
L
pictured on Fig. 3, passes above all the poles of

h() at height > 0,
and can be closed in the lower half-plane where e
it
behaves conveniently as to cancel the
integral at innity there. Note that although the requirement > 0 is emphasized in the book
(p. 139), I still have to understand why being above all the poles is not enough. And as we
shall see very soon, > 0 is the key to the choice of the right part of the complex plane.
Lets compute from the Maxwell-Vlasov Eqs. (1,2) the time evolution of the system con-
sidering,
F(x, v, t) = n
0
f
0
(v) + F
1
(v, t)e
ikx
, (12)
E(x, t) = E
1
(t)e
ikx
, (13)
assuming F
1
, E
1
are rst order quantities, and F
1
(v, t = 0)e
ikx
is the perturbation initially
applied. The linearized Vlasov equation reads,
F
1
(v, t)
t
+ ikvF
1
(v, t)
en
0
m
E
1
(t)f

0
(v) = 0. (14)
If we multiply by e
it
and take the integral from t = 0 to +, an integration by part on the
time derivative term gives,
_

0
e
it
F
1
(v, t)
t
dt =
_
e
it
F
1
(v, t)

0
i
_

0
e
it
F
1
(v, t)dt
= F
1
(v, 0) i

F
1
(v, ), (15)
where lim
t
e
it
F
1
(v, t) = 0 has been assumed. On the one hand, the very existence of the
Laplace transform of F
1
(v, ) =
_
F
1
(v, t)e
it
dt implies it. On the other hand, a important
conclusion of the paper is that for large times, F
1
(v, t) e
ikvt
(see [1] p. 452, and Plasma
5

/ k
/ k
v
/ k
Figure 4: Forbidden option for the contour. Continuity is preserved, but the contour lies above
the pole, in contradiction with the Laplace prescription.
Talk 5). This point is discussed neither in the book, nor in the original paper. Using Eqs.
(15,14) then gives,
(ikv i)

F
1
(v, )
en
0
m

E
1
()f

0
(v) = F
1
(v, 0), (16)
where F
1
(v, 0) now acts like a source term at the right-hand-side. A few more manipulations
exploiting Poissons equation (2) give,

E
1
() =
1
(k, )
4e
k
2
_

F
1
(v, 0)dv
v /k
, (17)
where (k, ) is identical to Eq. (5). The time dependant electric eld given by the inversion
formula (11) is,
E
1
(t) =
_
C
L
e
it

E
1
()d =
_
C
L
e
it
(k, )
_
4e
k
2
_

F
1
(v, 0)dv
v /k
_
d. (18)
In contradistinction with Eq. (5) where the contour issue is puzzling, the Laplace technique
used here is clear: The v-integration in (k, ) does go along the real axis, and the -integration
is performed at xed Im() = > 0. It means that in Eq. (18), which computes a physical
quantity, the dielectric function (k, ) is calculated with above the real v-axis.
That answers the question we had: the physically meaningful half-plane we were wondering
about after Eq. (5) is the upper one. The kind of contour pictured on Fig. 4 is thus
forbidden.
Incidentally, what are the poles of the integrand in Eq. (18)? For normal, smooth initial
excitations F
1
(v, 0), the term between brackets wont have poles, so that our poles
1
, . . . ,
n
are eventually the zeros of (k, ). The -integration of Eq. (18) on the closed contour C
L
will thus give, with
j
=
r,j
+ i
i
,
E
1
(t) = 2i
n

j=1
Res(j)
n

j=1
A
j
exp(i
j
t) =
n

j=1
A
j
exp(i
r,j
t)e

j
t
, (19)
which for large times will be governed by the largest
j
. Therefore, the Laplace transform
approach cannot spare us the resolution of (k, ) = 0, as these zeros are the building blocks
of the temporal response of the system.
6
3 Resolution for small damping
We suppose small damping, that is || |
r
|, and Taylor expand Eq. (6),
(k,
r
+ i) =
r
(k,
r
) + i

r

=0
+ i
_

i
(k,
r
) + i

i

=0
_
=
r
(k,
r
) + i
i
(k,
r
) +
_
i

r

r
_
=0
= (k,
r
) + i

r

=0
+ o(), (20)
where the o() (negligible with respect to ), comes from the fact that
i
( = 0) = 0 (no
damping, no dissipation, no imaginary dielectric function).
The rst term (k,
r
) is given by Eq. (6) setting =
r
, or taking the limit of (k,
r
+i)
for 0
+
. The part of the integration along the real axis for v [,
r
/k ] [
r
/k +
, +] gives the so-called Cauchy Principal Part, denoted P here. The part corresponding
to the semi-circle (see Fig. 2 middle) gives the semi-residue for v =
r
/k. An alternative way
of deriving this result, considering the limit 0
+
, is reported in Appendix A. We thus get,
(k,
r
) = 1

2
p
k
2
_
P
_
f

0
v
r
/k
dv + if

0
(
r
/k)
_
. (21)
This result allows to compute
r
/
r
in Eq. (20), which eventually gives,
(k, ) = 1

2
p
k
2
P
_
f

0
v
r
/k
dv i

2
p
k
2
_
f

0
(
r
/k) +

r
P
_
f

0
v
r
/k
dv
_
. (22)
Equating the real part to zero yields,

2
p
k
2
P
_
f

0
v
r
/k
dv = 1, (23)
which was the result obtained by Vlasov in the rst place. Cancelling the imaginary part gives
directly the damping rate,
=
f

0
(
r
/k)

r
P
_
f

0
vr/k
dv
. (24)
Eqs. (23, 24) formally solve the problem in terms of the distribution function. A rst order
evaluation of P k
2
/
2
r
(see Eq. (27) below), gives

r
=
p
, and then (25)

p
=

2

2
p
k
2
f

0
(
p
/k).
The rate has the sign of f

0
(
p
/k). That means that if f
0
decreases for v =
p
/k, the
waves is damped because < 0. But if f
0
increases for v =
p
/k, we have > 0 and the wave
can actually grow.
7
One could argue we started initially assuming positive, and nd it can be negative here.
It is not a problem for the following reason: Eq. (22) we found assuming > 0 is continuous
at = 0. It must therefore be identical to the integration on the Landau contour on both
sides of the real axis. We can therefore condently solve it regardless of the sign of . In other
words, thanks to the Landau contour, we can compute the result as if was positive, and then
dont care about the sign.
Historically, Vlasov rst ran into Eq. (5). He escaped the problem posed by the pole on
the real axis by considering only the P of the quadrature. He did so apparently without much
foundation, which Landau denounced without mercy in [1]. We understand from the analysis
above that doing so, he missed the imaginary part which would have led to the Vlasov
damping.
4 Maxwellian distribution
Lets nally consider a 1D Maxwellian distribution,
f
0
(v) =
1
_
2k
B
T/m
e
mv
2
/2k
B
T
. (26)
For phase velocities
r
/k much larger than the thermal velocity V
th
=
_
k
B
T/m, we can
expand the denominator in powers of kv/
r
, since that quantity is small where the numerator
is relevant. We thus have,
P
_
f

0
v
r
/k
dv =
k

r
_
f

0
_
1 +
kv

r
+
k
2
v
2

2
r
+
k
3
v
3

3
r
+
_
dv
=
k
2

2
r
+ 3
k
B
T
m
k
4

4
r
+ (27)
For small k, namely kV
th
/
r
kV
th
/
p
1, Eq. (23) now gives,

2
r
=
2
p
(1 + 3k
2

2
D
), with
D
=
_
k
B
T/m

p
. (28)
We nally (phew!) use Eq. (24) to extract the damping rate. On the one hand, we compute
the derivative of the P with respect to
r
using Eq. (27), and then simply set
r
=
p
in the
result. On the other hand, we set
r
=
p
in f

0
to nd
5
=
p
_
/8
k
3

3
D
exp
_

1
2k
2

2
D
_
. (29)
Fluid theory just gives the real part of the frequency, namely Eq. (28), so that Landau
damping is a purely kinetic eect.
5
Some authors insert the full expression of r from Eq. (28), yielding 1/2k
2

2
D
3/2 in the argument of
the exponential.
8
4 2 2 4 6 8 10
v
0.2
0.4
0.6
0.8
1.0
G1
4 2 2 4 6 8 10
v
1
2
3
4
5
G2
Figure 5: Functions G
1
and G
1
involved in Eq. (32). For small /k, G
1
is almost 1 everywhere,
except for v =
r
/k, where it is 0. G
2
peaks at v =
r
/k and tends to 0 elsewhere, while its
integral is always , like a Dirac function. Parameters are
r
/k = 4, /k = 0.2 for G
1
, and
/k = 0.2, 0.5, 1 for G
2
.
Appendix A
Lets derive,
_
L
f

0
v
r
/k
dv = P
_
f

0
v
r
/k
dv + if

0
(
r
/k), (30)
used for Eq. (21), without using the residue theorem. For =
r
+i with > 0, integration
along the Landau contour is equivalent to an integration along the real axis. Lets thus assume
> 0 and compute,
I = lim
0
+
_

0
v (
r
+ i)/k
dv. (31)
We multiply the numerator and the denominator of the integrand by (v
r
/k) + i/k,
which is the complex conjugate of the denominator. We get an expression with a purely real,
non singular denominator, and clearly separated real and imaginary parts,
I = lim
0
+
_

(v
r
/k)
2
(v
r
/k)
2
+ (/k)
2
. .
G
1
f

0
(v
r
/k)
dv + i
_

/k
(v
r
/k)
2
+ (/k)
2
. .
G
2
f

0
dv. (32)
Regarding the real part, the factor G
1
of the integrand is 0 for v =
r
/k, and 1 for
/k |v
r
/k|. It tends to the P
_
f

0
/(v
r
/k) for small /k (see Fig. 5). The factor G
2
of the integrand of the imaginary part departs from 0 only for v
r
/k. But its integral is
always . For small /k, the quadrature thus tends to f

0
(
r
/k), and we are back to (30)
6
.
6
The limit of i with 0
+
is sometimes written i0. The identity
lim
0
+
Z

h(x)
x a i
dx
Z

h(x)
x a i0
dx = P
Z

h(x)
x a
dx + ih(a),
can be referred to as the Plemelj Formula in the literature. For 0

, the imaginary part above is ih(a).

9
This calculation is consistent with the Landau contour integration only for 0
+
. This
is because in such case, the real axis along which we perform the integration (32) coincide
with the Laundau contour. If we were to compute Eq. (32) for 0

, we would nd the
opposite imaginary part, just because in this case, the real axis no longer ts the Landau
contour. The latter, instead, is deformed and keeps passing below the pole, precisely to avoid
the discontinuity.
References
[1] L.D. Landau, J. Phys. (U.S.S.R.) 10, 25 (1946).
[2] A. Vlasov, J. Phys. 9, 25 (1945).
[3] L.D. Landau and E.M. Lifshitz, Course of Theoretical Physics, Physical Kinetic.
10